Master's Thesis Matthias Greinert


Optimization of the operation of heat pump systems

In this master thesis the possibilities of optimizing the operational mode of a heat
pump system using a metaheuristic are tested. The aim of the optimization is a
reduction of the consumed primary energy, while ensuring an acceptable level of
thermal comfort. This is a combinatorial optimization problem. Usually it is impossible
to calculate all solutions belonging to a combinatorial optimization problem,
due to extensive computation time. An approximate solution is being searched
instead of the global minimum. Metaheuristics are a reasonable tool for achieving
this. The heat pump system is modelled within a simulation environment and the
metaheuristic optimizes the simulation results in an iterative manner. This leads to
a coupled simulation and optimization process. A short overview of known metaheuristics
and their corresponding behavior is given. The properties of the heat
pump system are analyzed and a suitable metaheuristic for solving the optimization
problem at hand is derived. It is explained how the metaheuristic can be adjusted
to match the specific structure of the optimization problem. During this, the
used algorithms for optimization, the handling of a multi-objective optimization and
the design of the objective function are focused. The metaheuristic is used to optimize
a heat-driven solution, which is also applied as an initial solution to the optimization
problem. A comparison of the optimized and the heat-driven solution
illustrated the potential for optimization. A genetic algorithm is implemented to
benchmark the quality of the metaheuristic developed in this master thesis.